Chasing down a 100Ω short to ground in 820-01521-A

[sekidata]

This 820-01521 2019 Air board had liquid damage in several places, was missing PPBUS, and had shorts on both CD3215s. I replaced all corroded components I could find, including both CD3215s. Now I'm left with PP5V_G3S which shows 80-100 Ω to ground (0.07 in reverse diode mode). No shorts on any other rail I could find.

I already lifted every chip powered by PP5V_G3S but haven't found the culprit for the short. I can't inject 5V into 100 Ω (not enough current flow). How can I find the short?

 

[2informaticos]

Did you check when lifted U3100/200 chips?
Tested disconnecting fan?

Removed all L6703, R7100, U7210/30/90, U7410/30, R8103/44/47, R8444/45, U8500, L8570?
If you lift L7670, the low value remains on pad 1 for sure?

 

[sekidata]

Yes, I have both CD3215s off the board again.
Yes, when I lift L7670, the short is on pad 1 but not on pad 2.
Yes, I tried lifting all resistors and inductors (one by one) that connect from PP5V_G3S to subnets: L8570, L6703, R7100, R8103, R8144, R8147, R8444, and R8445.
And I lifted U7210/30/90, U7410/30, U8500, and U7650.

 

[2informaticos]

Then you have problems with a capacitor connected on that line.
You can try applying 5V there: the culprit can start droping its resistance under voltage.
IR camera is recommended in such case.

 

[sekidata]

No thermal imaging here, sorry.

I just replaced U8110 and my short changed from 100 to 400 Ω. Doesn't make a lot of sense to me. Could there be a short inside the board that reacts to the temperature changes during soldering?

 

[2informaticos]

I usually found bad capcitor, with that behavior.

 

[sekidata]

After lifting every capacitor on this rail, no luck. It will remain a mystery.

 

[2informaticos]

Post diode mode readings in both ways to ground.

 

[sekidata]

Diode mode red-to-ground: 0.246, black-to-ground: 0.250
Resistance: 305 Ω

 

[2informaticos]

Still something bad there...

 

[sekidata]

I just placed all components back on the board and it looks like PP5V_G3S is at 5.1V, even with the short. No startup, though.

 

[2informaticos]

That is not really a short; but not expected value there however.

So you get stable 5V_G3S now?
Tested DFU revive?

 

[sekidata]

Yes, PP5V_G3S is stable now. DFU Revive ends in error -1.

 

[2informaticos]

Can you post picture of the screenshot?

3V3_G3S and 1V8_G3S are good too?
Any voltage on SSD power rails?

The data is probably lost, ask the client if accept DFU restore.
BTW, keep battery connected, when try DFU.

 

[sekidata]

Data is irrelevant so I did a Restore. Note that I had a hard time getting the machine into DFU mode, so I used a blank SoC ROM chip to force DFU.

I have:
PP3V3_G3S: 3.2V
PP1V8_G3S: 1.8V
PP0V9_SSD: 0.9V
PP1V8_SSD: 1.8V
PP2V7_NAND: 2.5V

 

[2informaticos]

Did you connect all the things in housing to MLB?
Be sure to use good battery and trackpad.

 

[sekidata]

No, just the logic board, USB, and battery. Is that really necessary?

 

[sekidata]

Ok, with all peripherals connected, the DFU Restore went through to the padlock. But no video on the target machine. No startup. Assume bad CPU?

 

[2informaticos]

Do you get any voltage later than G3S state?

Compare diode mode readings on thebig coils with good board.
Also test the board disconnecting audio flex.